www.nature.
com/scientificreports
OPEN
Received: 7 July 2017
Accepted: 28 November 2017
Published: xx xx xxxx
ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4
Remote control of resistive
switching in TiO2 based resistive
random access memory device
Dwipak Prasad Sahu & S. Narayana Jammalamadaka   
We report on the magnetic field control of a bipolar resistive switching in Ag/TiO2/FTO based resistive
random access memory device through I–V characteristics. Essentially, in the presence of magnetic
field and in the low resistance state, an abrupt change in the resistance of the device demands higher
voltage, hinting that residual Lorentz force plays a significant role in controlling the resistance state.
Endurance characteristics of the device infer that there is no degradation of the device even after
repeated cycling, which ensures that the switching of resistance between ‘off’ and ‘on’ states is
reproducible, reversible and controllable. Magnetic field control of ‘on’ and ‘off’ states in endurance
characteristics suggest that this device can be controlled in a remote way for multi-bit data storage.
Non-volatile memory (NVM) technology indeed requires intensive research as the conventional silicon (Si) based
memories are approaching to their scaling limits. In particular, among various promising modern NVM technol-
ogies, resistive random access memory (RRAM) has attracted a great deal of scientific and technological interest
owing to its easy fabrication, high density and promising performances. RRAM devices would work based on
the resistive switching (RS) phenomena in which the resistance state can be altered between high resistance state
(HRS) and low resistance state (LRS) by controlling applied voltage1,2. If both the positive and negative voltages
are essential to switch resistance state between HRS and LRS, such RS can be termed as bipolar resistive switching
(BRS)2. On the other hand, if only one polarity of the voltage is sufficient to switch between HRS and LRS, such
RS can be termed as unipolar RS (URS). Controlling RS with voltage has been demonstrated by various authors
in oxides3, nitrides4 and organic materials5.
On top of that controlling the RS effect with thickness variation has been demonstrated previously. An
increase in the reset voltage has been observed by Wang et al., upon increase in the thickness of ZnMn2O4 in Ag/
ZnMn2O4/p+-Si based RRAM devices6. On the other hand, Zhu et al., have reported thickness dependent bipo-
lar resistive switching behavior pertinent to NiOx films, where they observed the clockwise (for low thickness)
and anticlockwise (for high thickness) I–V characteristics respectively7. Yet in another study, Kang et al., have
reported that upon increase in the thickness of ZnO in Al/ZnO/Al layered RS memory device, indeed there is
an enhancement in the crystallinity of ZnO layer, the concentration of oxygen related defects and set voltages8.
On the other hand, sharath et al., have reported the effect of HfO2 layer thickness on the forming voltage. It has
been observed that forming voltage increases linearly with HfO2 layer thickness9. Ito et al., have reported on the
oxide thickness dependence of resistive switching characteristics for Ni/HfOx/Pt resistive random access mem-
ory device. Their observation has indicated that a clear dependence of switching voltages for the set and reset
processes on oxide thickness10. Enhanced values of forming voltages have been observed by Inoue et al., upon
increase in thickness of Fe2O3 in the Pt/Fe2O3/Pt based RRAM devices11.
Despite the control of RS effect with voltage and thickness of oxide film, the underlying mechanism for the RS
effect has been under intense debate. However, two successful models those have been proposed to explain RS
mechanism are: (a) Interface model with a redox reaction at the interface between the electrode and the insulating
layer (such as modification of Schottky barrier height or width)12 and (b) bulk effect or filament model (such as
formation of local conductive filaments)13.
The crucial parameter to control transport properties/switching mechanism in aforementioned devices is
voltage, hinting that only one degree of freedom exists in order to control the switching mechanism. However,
recent technologies demand more than one degree of freedom to control switching mechanism. Hence, there is
a quest for the manipulation of switching effect with external parameters such as magnetic field, light14,15 and
Magnetic Materials and Device Physics Laboratory, Department of Physics, Indian Institute of Technology
Hyderabad, Hyderabad, 502 285, India. Correspondence and requests for materials should be addressed to S.N.J.
(email: surya@iith.ac.in)
1
www.nature.com/scientificreports/

Figure 1. (a) Grazing angle XRD pattern of TiO2/FTO thin film. It is evident from the figure that in addition to
anatase TiO2 peaks, peaks from substrate are also present (b) FE-SEM image of TiO2 thin film. It is clear from
the graph that the grain size is 50–60 nm.

temperature16. Among them, magnetic fields give an opportunity to control transport/switching mechanism in a
remote way. Hence, in the present manuscript, we put our efforts in controlling the transport/switching properties
of Ag/TiO2/FTO based RRAM device in a remote way using magnetic fields. Earlier, magnetic field control of RS
has been well demonstrated on devices which are doped with magnetic elements17,18. Bai Sun et al.14 showed a
systematic increase in set and reset voltage of Ag/[BiFeO3/γ-Fe2O3]/FTO device with increasing field strength,
which has been explained on the basis of coupling between magnetism and ferroelectricity of BFO. Yet in another
study, a delay in transition from HRS to LRS in presence of magnetic field has been well reported by Wang et al.19
in Si–SiO2–MgO device. The resistive switching behavior has been most widely observed in a variety of binary
transition metal oxides20–22 such as NiO, ZnO, TiO2, Nb2O5 and ZrO2.
Recent reports on anatase TiO2 in the form of thin film has demonstrated significant magnetic properties23.
However, magnetic field control of RS in anatase TiO2 has not been explored until now. Essentially the magnetic
field may create Lorentz force on charge carriers and influence I–V characteristics. Hence, in the present man-
uscript, we would like to utilize this intriguing phenomena to control RS in anatase TiO2 based RRAM devices.
Salient features of the present manuscript are (a) controlling the abrupt change of resistance in the low resistance
state with a magnetic field (b) tuning the endurance characteristics with magnetic fields (c) attaining multi-bit
storage with voltage control and altering it with the magnetic fields.

Results and Discussion

Figure 1(a) shows the grazing angle x – ray diffraction pattern obtained from TiO2 deposited on FTO substrate
(TiO2/FTO) after annealing at 400 °C for 1 hour. The diffraction peaks at 25.3° and 47.9° are assigned to (101) and
(200) planes respectively, hinting that TiO2 is in the anatase phase. All peaks are in line with respect to the stand-
ard spectrum pertinent to anatase TiO2 (JCPDS no. 84-1286). Apart from TiO2 peaks, we also observed peaks
from the substrate FTO, which is evident from Fig. 1(a). Figure 1(b) shows the FE-SEM image of TiO2 film. It is
evident that the grain size is around 50–60 nm. Composition analysis through energy dispersive x – ray spectros-
copy (EDX) indicated that Ti and O are in 1:2 phase.
Figure 2(a) shows the schematic diagram of the experimental setup that we used to perform measurement on
TiO2 based RRAM device. Figure 2(b) and (c) depicts the 3D C – AFM images from TiO2 thin films when the
applied bias voltages are 0 V and −8 V respectively. More details about the Fig. 2(a–c) are discussed in methods
section.
Figures 3(a) and (b) reveals I–V characteristics of Ag/TiO2/FTO device in linear and logarithmic scales
respectively. From Fig. 3(a) it is evident that initially, the device is at HRS. By sweeping the voltage from 0 to −5 V
with a current compliance of 40 mA (to avoid dielectric breakdown), indeed there is a smooth decrease of current
until −3.4 V, above which a sharp jump is evident and device switches to LRS. In the reverse cycle from −5 V to
+5 V, the device stayed in LRS and changes to HRS state in the cycle between +5 V to 0 V. This means that in order
to switch the device from HRS to LRS (SET switching) (at −3.7 V) or LRS to HRS (RESET switching) (at +3.5 V),
we need to apply two different polarities of voltage, which is typical for a BRS. The arrow marks with numbers on
the graph shows the sequence that is followed while recording I–V characteristics.
In order to understand the mechanism of resistive switching in our device, we used different conduction
models to fit the I–V data. As shown in Fig. 3(a) inset (i), at low voltage region, log (I) vs. log (V) shows ohmic
conduction behavior (I α V) with a slope of about ~1.1. This indeed followed by a quadratic nature at high voltage
regime, which corresponds to Child’s square law (I α V2) with a slope of nearly ~2.0. This type of switching behav-
ior from current-voltage relationship can be explained by trap controlled space charge limited current conduc-
tion mechanism (SCLC) model where the dielectric layer is triggered by deficiency of oxygen. According to this
conduction mechanism, at low voltage region, the thermally generated free charge carrier density inside the film

ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4 2

www.nature.com/scientificreports/

Figure 2. (a) Schematic diagram of the measurement performed on TiO2 based RRAM device where the
magnetic field is applied perpendicular to the direction of the current. Keithley 2400 is used to source the
voltage and sense the current respectively (b) & (c) 3D conducting surface atomic force (C - AFM) images
scanned in an area of 1 × 1 µm2 by applying a voltage bias of 0 and −8 V respectively.

Figure 3. (a) The I–V characteristics of the Ag/TiO2/FTO device in the range 0 V → −5V → 0 V → +5 V → 0 V.
Inset indicates fitting results (i) LRS follows space charge limited conduction mechanism and (ii) HRS follows
Schottky emission mechanism (ln I~V1/2) (b) I~V curve plotted in semi-logarithmic scale.

ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4 3

www.nature.com/scientificreports/

Figure 4.  I–V curve of Ag/TiO2/FTO device at different magnetic fields. It is evident from the figure that the
abrupt voltage (VA) shifts to higher voltages with magnetic field due to residual Lorentz force. Inset shows of
variation of abrupt voltage (VA) with magnetic field.

dominates the injected carrier density resulting in an ohmic behavior. When the applied voltage (V) is smaller
than the minimum voltage (Vtr = voltage required to transit to space charge limited region), the injected carriers
redistribute themselves internally due to excess dielectric relaxation time in order to maintain charge neutrality.
Concurrently, the probability for injected carriers which can travel across the insulating film is zero. The outset
of departure from ohm’s law takes place only when applied voltage reaches Vtr. Thus, with increasing the applied
voltage, the density of injected carriers exceeds the thermally generated carriers to such a value that the Fermi
level moves up above the electron trapping level. When all traps are filled, the injected carriers are free to move in
the film and excess charges build up in the film, thereby making a transition to SCL conduction, which explains
the LRS behavior. Therefore, the generation of conducting path is favored by formation of electron traps and may
be due to oxygen vacancies in the TiO2 layer results in the change of resistance state pertinent to the device.
On the other hand, in case of HRS, the current behavior follows Schottky emission (SE) mechanism which is
verified by ln (I) α V1/2 graph as shown in the Fig. 3(a) inset (ii)). The fitting results indicate that due to the ther-
mionic emission between the interface of switching layer and an electrode, electrons cannot climb the potential
barrier, which leads less conduction.
Now we discuss about the effect of the magnetic field on the BRS in TiO2 based RRAM device. As we men-
tioned in methods section, we applied the external magnetic field parallel to the surface of the film (perpendicular
to the current path). From Fig. 3 it is evident that in LRS state there is a sharp change of current from 15 mA
to 40 mA around 3.5 V. We define this as an abrupt voltage (VA). It is to note that as we apply large amount of
magnetic field, high value of VA is required, which manifests the control of transport properties are in a remote
way. The evolution related with shift of VA in a remote way is shown in linear plots (Fig. 4). We also performed
the measurements when magnetic field is parallel to the current direction. We do not see any shift of VA with H
(graph is not shown here as there is no shift). Inset of Fig. 4 depicts the variation of VA with the externally applied
transverse magnetic field. The probable mechanism for the variation of VA with the transverse magnetic field can
be explained as follows.
Essentially, when a transverse magnetic field is applied with respect to electric field, a Hall field is induced in
the device, thereby making the total electric field as a vector sum of applied field and an induced field. Also, under
the transverse field, the charge carriers experience a Lorentz force which is thought to be completely cancelled
by the induced Hall field only if carriers move with an average drift velocity. If the drift velocity is more than the
average drift velocity, the Lorentz force dominates over the force of Hall field resulting in a residual Lorentz force
for charge carriers. Mathematically the Lorentz force can be expressed as
F = q(E + v × μ0H )

where q is charge of electron, E applied electric field, v is velocity of electron μ0 is permeability in free space and
H externally applied magnetic field. As can be seen from Fig. 2(a), the direction of electric field is from top to
bottom and the magnetic field is from left to right. By considering the negative charge on electron the Lorentz
force on electron should be out of the paper. Ultimately, a force would influence the motion of electrons due to
simultaneous application of electric and magnetic field perpendicular to each other. As a result of this force,
electrons would attain cycloid motion, which demands larger value of VA with the magnetic field. Fundamentally,
as the magnetic field increases, the force on the electrons increases and hence an increase in VA. From this experi-
ment we realize that the residual Lorentz force effect exists only when H is perpendicular to current direction and
influences the I–V characteristics of the TiO2 based RRAM device.

ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4 4

www.nature.com/scientificreports/

Figure 5.  Schematic diagram to explain the bipolar switching mechanism in the absence of magnetic field
which corresponds to formation and rupturing of conductive filaments which may be due to electro-migration
of oxygen ions.

The conceivable mechanism for the observed RS behavior and filamentary formation can be explained as fol-
lows. Figure 5 explains about forming and rupturing of the filaments in Ag/TiO2/FTO RRAM device. Essentially,
by the application of negative voltage, electro-migration of oxygen ions takes place and may create oxygen vacan-
cies. This indeed leads to formation of filaments between Ag and FTO electrodes, which results in the LRS state.
In contrast, for the applied +ve voltage, due to the repulsion between oxygen vacancies and positive charges,
rupturing of the filaments takes place, which leads to HRS state.
Earlier, the shift of voltage with the applied magnetic field has been demonstrated in GaAs devices24 and
correlated such an intriguing phenomena to a residual Lorentz force, shift of Landau level of electrons in con-
duction band, electron mobility and thermal & impact ionization of electrons25,26. Yet in another work, the effect
of magnetic field on switching behavior of silicon device also showed the suppression of LRS state confirming
that magnetic fields can influence the LRS state19. Earlier conductive bridge resistive random access memory
(CBRRAM) cell using Ag doped polymer electrolyte between Pt electrodes has well been demonstrated27. We
rule out this possibility in our current device due to the following reason. (a) we have done extensive analysis of
FESEM between two contact Ag pads. We do not see the presence of Ag anywhere between two contact pads upto
the depth profile of 1 μm (b) the gap between two Ag contacts is ~1.2 cm.
Endurance characteristics of TiO2 RRAM device are shown in the inset of Fig. 6. Memory window of the cell
is calculated using the formula (ROFF − RON)/RON ≈ ROFF/RON and is found to be 10. Indeed, this window is huge,
which essentially makes the device suitable to distinguish the information between 1 and 0. It is also evident that
there is no degradation of the device even after repeated cycling which ensure the switching between off and on
states is reproducible, reversible and controllable. In order to get more insights on durability of the device, we
performed endurance characteristics in the presence of magnetic field using Keithley 4200 semiconductor char-
acterization system. For this purpose, a current compliance of 10 mA and a pulse width of 10 µs with set/reset
voltage of −4 V/+2 V is applied to the device. From Fig. 6 it is apparent that the resistance values in both high and
low resistance states increase with magnetic field. Such an increase in resistance values with magnetic field can be
attributed to Lorentz force as we mentioned earlier.
The endurance characteristics of the device also tested by the application of four different voltage pulses of
−7V, −6V, −5V, −4V and +2 V with a pulse width of 50 µs (Fig. 7). Here, negative voltages are applied to the
device to set it into different on states and the reset was achieved by applying positive voltage pulse of + 2 V. For
example, at −6 V, the device is in low resistance state. It is evident from the inset (a) of Fig. 7 that, there is a sys-
tematic increase in the resistance values of the device at various magnetic fields. We also tried to plot the change
in resistance (ΔR/R) with respect to zero field value as shown in the inset (b) of Fig. 7. It is evident that nearly 5%
change in the resistance value is evident by the application of 2300 Oe, which demonstrates that the resistance
state can precisely be tuned in a remote way. Indeed, we could tune resistance state for different on states (at −7 V,
−6 V, −5 V and −4 V) as well as for off state (+2 V) as shown in Fig. 7. These results are reproducible and con-
sistent, hinting that tuning of multilevel resistive switching of Ag/TiO2/FTO device with the magnetic field in a
remote way that promises future application in multi-bit data storage technology.

Methods
The phase of TiO2 was confirmed by (Bruker Discover D8) X-ray diffraction (XRD) with Cu - Kα radiation
(λ = 1.54 Å). Microstructural studies were performed using the field emission scanning electron microscope
(Zeiss ultra 55 FE-SEM). Composition of the films was determined using the energy dispersive x- ray spectros-
copy (EDX) attached with the FE-SEM.

ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4 5

www.nature.com/scientificreports/

Figure 6.  Endurance characteristics of resistive switching for different magnetic fields in the range 0–2300 Oe.
It is evident that with magnetic fields, there is an enhancement in the resistance values in LRS and HRS. Inset
shows endurance characteristics with a memory window about 10.

Figure 7.  Multilevel resistive switching in Ag/TiO2/FTO device at different voltage pulses (−7 V, −6 V, −5 V,
−4 V and +2 V) with a width of 50 µs under various magnetic fields in the range 0–2300 Oe.

RRAM devices were fabricated on Fluorine doped Tin Oxide (FTO) substrate in order to elucidate the resis-
tive switching behavior in Ag/TiO2/FTO device. Initially, TiO2 paste (obtained from Solaronix) was used to pre-
pare a thin film on FTO substrate using a drop casting method and subsequently the film was annealed for 1
hr at 400 °C. TiO2 layer thickness is around 5 µm. On top of this annealed TiO2 films, contacts were given with
silver epoxy (~0.3 mm). The distance between two silver (Ag) contacts is nearly 1.2 cm. Here we used Ag as top
electrode and FTO as bottom electrode. Ag was used as top electrode because it provides very good conduction.
In addition, Ag has commonly been used as an inert electrode material due to its anti-oxidation property. FTO
was used as bottom electrode as the work function of FTO and TiO2 are close to each other with a difference of
0.4–0.6 eV, we expect the flow of electrons would be easy from FTO to TiO2 with a little amount of potential. The
resistance of FTO is approximately 10 Ω. Two probe method was used to perform I–V characteristics of Ag/TiO2/
FTO device at room temperature using a Keithley 2400 meter with a compliance limit of 40 mA. Subsequently,
the effect of the magnetic field on the switching behavior was carried out. I–V characteristic of the Ag/TiO2/FTO
device in the presence of magnetic field was performed in two configurations (a) without magnetic field (H = 0)
and (b) with magnetic field (H ≠ 0) (Fig. 2(a)). In the configuration where H ≠ 0, an electromagnet (Walker sci-
entific) was used to generate the magnetic field in the range of 0 to 3000 Oe (Fig. 2(a)). The direction of magnetic
field is perpendicular to the motion of electrons in the device (CPP configuration). Keithley 4200 (semiconductor
characterization system) was used to perform the endurance characteristics in the presence of magnetic field.
Essentialy, the endurance characteristics of the device tested by the application of four different voltage pulses
of −7 V, −6 V, −5 V, −4 V and +2 V with a pulse width of 50 µs at various magnetic fields. In order to confirm

ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4 6

www.nature.com/scientificreports/

the role of conductive filaments in the device, we have characterized the device with conductive atomic force
microscopy (CAFM). In C-AFM study, the metallic AFM tip behaves as a nano-sized electrode in the switching
of the Metal/Insulator/Semiconductor structure. In this method, a voltage was applied between the Pt/Ir coated
antimony n - doped Si tip (Commercial Code SCM-PIC) that serves as top electrode and FTO as bottom elec-
trode. The curvature radius, oscillation frequency and cantilever spring constant of the tip are 20 nm, 13 kHz and
0.2 N/m respectively. The experiment was carried out at room temperature and atmospheric pressure without
applying any current compliance. We applied bias voltages of 0 V and −8 V in two different experiments on an
area of 1.0 × 1.0 µm2 at 1.01 Hz in contact mode. When the voltage is 0 V, we do not see any conducting channels
in the film (Fig. 2(b)). However, when the bias voltage is −8 V, we do see conducting channels which may be due
to oxygen vacancies as shown in Fig. 2(c). This indicates the presence of conducting channels that are distributed
across the scanned area. The current mapping image confirms the widely accepted filament model for the resistive
switching in TiO2.

Summary.  In summary, we demonstrated the stable bipolar resistive switching in Ag/TiO2/FTO device and
also explored the effect of the magnetic field on the resistance switching behavior. We could tune the abrupt
voltage (VA) in LRS state remotely and explained on the basis of Lorentz force effect. Endurance characteristics
inferred an increase in resistance values with the magnetic field. Tuning the resistance state for different on states
(at −7 V, −6 V, −5V and −4 V) as well as for off state (+2 V) with magnetic field reveal that resistance states (both
LRS and HRS) can be tuned in a remote way. This indeed hint that the present results may be helpful in future
RRAM based devices those would operate with magnetic fields.

References
1. Waser, R., Dittmann, R., Staikov, G. & Szot, K. Redox‐based resistive switching memories–nanoionic mechanisms, prospects, and
challenges. Advanced materials 21, 2632–2663 (2009).
2. Lee, J. S., Lee, S. & Noh, T. W. Resistive switching phenomena: A review of statistical physics approaches. Applied Physics Reviews 2,
031303 (2015).
3. Rozenberg, M. J. et al. Mechanism for bipolar resistive switching in transition-metal oxides. Physical Review B 81, 115101 (2010).
4. Kim, H. D., An, H. M., Lee, E. B. & Kim, T. G. Stable bipolar resistive switching characteristics and resistive switching mechanisms
observed in aluminum nitride-based ReRAM devices. IEEE Transactions on Electron Devices 58, 3566–3573 (2011).
5. Cho, B., Song, S., Ji, Y., Kim, T. W. & Lee, T. Organic resistive memory devices: performance enhancement, integration, and advanced
architectures. Advanced Functional Materials 21, 2806–2829 (2011).
6. Zhang, Y. P. et al. Effect of ZnMn2O4 thickness on its resistive switching characteristics. Indian Journal of Engineering &Material
Sciences 21, 563–566 (2014).
7. Zhu, H. X. et al. Thickness-Dependent Bipolar Resistive Switching Behaviors of NiOx Films. In. Materials Science Forum 847, 131
(2016).
8. Kang, Y. H., Choi, J. H., Lee, T. I., Lee, W. & Myoung, J. M. Thickness dependence of the resistive switching behavior of nonvolatile
memory device structures based on undoped ZnO films. Solid State Communications 151, 1739–1742 (2011).
9. Sharath, S. U. et al. Thickness independent reduced forming voltage in oxygen engineered HfO2 based resistive switching memories.
Applied Physics Letters 105, 073505 (2014).
10. Ito, D., Hamada, Y., Otsuka, S., Shimizu, T. & Shingubara, S. Oxide thickness dependence of resistive switching characteristics for
Ni/HfOx/Pt resistive random access memory device. Japanese Journal of Applied Physics 54, 06FH11 (2015).
11. Inoue, I. H., Yasuda, S., Akinaga, H. & Takagi, H. Nonpolar resistance switching of metal/binary-transition-metal oxides/metal
sandwiches: Homogeneous/inhomogeneous transition of current distribution. Physical Review B 77, 035105 (2008).
12. Nagashima, K., Yanagida, T., Oka, K. & Kawai, T. Unipolar resistive switching characteristics of room temperature grown SnO2 thin
films. Applied Physics Letters 94, 242902 (2009).
13. Kim, K. M. et al. Collective motion of conducting filaments in Pt/n‐type TiO2/p‐type NiO/Pt stacked resistance switching memory.
Advanced Functional Materials 21, 1587–1592 (2011).
14. Sun, B., Liu, Y., Zhao, W. & Chen, P. Magnetic-field and white-light controlled resistive switching behaviors in Ag/[BiFeO3/γ-Fe2O3]/
FTO device. RSC Advances 5, 13513–13518 (2015).
15. Li, H. et al. Light and magnetic field double modulation on the resistive switching behavior in BaTiO3/FeMn/BaTiO3 trilayer films.
Physics Letters A 381, 2127–2130 (2017).
16. Das, S., Majumdar, S. & Giri, S. Magnetic field, temperature, and time controlled manipulation of switching mechanism in NiO film:
Evidence of large magnetoconductance. The Journal of Physical Chemistry C 114, 6671–6675 (2010).
17. Chen, G. et al. Resistive Switching and Magnetic Modulation in Cobalt‐Doped ZnO. Advanced Materials 24, 3515–3520 (2012).
18. Krzysteczko, P., Reiss, G. & Thomas, A. Memristive switching of MgO based magnetic tunnel junctions. Applied Physics Letters 95,
112508 (2009).
19. Wang, J. et al. Magnetic field controllable nonvolatile resistive switching effect in silicon device. Applied Physics Letters 104, 243511
(2014).
20. Seo, S. et al. Electrode dependence of resistance switching in polycrystalline NiO films. Applied Physics Letters 87, 263507 (2005).
21. Chen, C., Pan, F., Wang, Z. S., Yang, J. & Zeng, F. Bipolar resistive switching with self-rectifying effects in Al/ZnO/Si structure.
Journal of Applied Physics 111, 013702 (2012).
22. Kim, K. M., Choi, B. J., Shin, Y. C., Choi, S. & Hwang, C. S. Anode-interface localized filamentary mechanism in resistive switching
of TiO2 thin films. Applied physics letters 91, 012907 (2007).
23. Hong, N. H., Sakai, J., Poirot, N. & Brizé, V. Room-temperature ferromagnetism observed in undoped semiconducting and
insulating oxide thin films. Physical Review B 73, 132404 (2006).
24. Aoki, K., Kondo, T. & Watanabe, T. Cross-over instability and chaos of hysteretic I–V curve during impurity avalanche breakdown
in n-GaAs under longitudinal magnetic field. Solid state communications 77, 91–94 (1991).
25. Tzeng, S. Y. T. & Tzeng, Y. Two-level model and magnetic field effects on the hysteresis in n–GaAs. Physical Review B 70, 085208
(2004).
26. Lee, F. S. & Cheng, Y. C. Magnetic-field effects on the hysteresis in semiconductors with an S-shaped negative differential
conductivity. Physical Review B 56, 6412 (1997).
27. Song, M. J., Kwon, K. H. & Park, J. G. Electro-Forming and Electro-Breaking of Nanoscale Ag Filaments for Conductive-Bridging
Random-Access Memory Cell using Ag-Doped Polymer-Electrolyte between Pt Electrodes. Scientific Reports 7, 3065 (2017).

ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4 7

www.nature.com/scientificreports/

Acknowledgements
We would like to acknowledge Indian Institute of Technology, Hyderabad for providing financial support. The
author Dwipak Prasad Sahu is thankful to the Department of Science and Technology, India (DST-INSPIRE) for
the award of junior research fellowship (JRF).

Author Contributions
S.N.J. and D.P.S. conceived and conducted the experiments. Both have analyzed the results and written the
manuscript.

Additional Information
Competing Interests: The authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

© The Author(s) 2017

ScIenTIfIc REPOrTS | 7: 17224 | DOI:10.1038/s41598-017-17607-4 8